Diagnosis of Respiratory Infections Using Syndromic Panels: A Ct-Based Approach Beyond Qualitative Detection

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors presented an interesting correlation between pathogen detection and pathogen dominance in codetections, in addition to two-year respiratory pathogen prevalence and seasonality.

Actually, it is documented experimentally that influenza virus dominates other pathogens in case of coinfection. Check these two articles (https://doi.org/10.1038/s43856-026-01504-x, https://doi.org/10.1038/s41467-024-53872-4)

 

The manuscript is interesting to general readers, however, needs to address these comments:

1. Abstract

• The first 2 sentences in the abstract need attention and rephrasing.
• line 18-20: please indicate that the study is retrospective and the years that it spans.

2. Methodology

Study Design and Sample Collection:

• Please mention the years of the data collection to give clear indication of the study period. It is not sufficient to say 2 years.
• This is assay is only validated for nasopharyngeal swabs according to the kit’s leaflet. So nasal swabs are not allowed to be used on this machine in diagnostic settings unless they are validated in the lab internally. Please comment on this.

Ct value extraction and semi-quantitative analysis

• (Lines 99-103) The delta Ct calculation is not quite clear. Please elaborate and clarify

Statistical analysis: why did they chose a median based comparison between delta Ct values over the mean delta Ct. This needs to be explained in the manuscript

3. Results;

• Table 1 is mentioned before figure 1 in the text but appears after it in the arrangement. Please rearrange.
• Supplementary figure 1 B can be arranged from highest to lowest to give a good idea about the prevalence.
• The same can be done for the table 1
• Ct values analysis:

The authors described different high and low Ct values with different pathogens. But there is no clear definition of high and low Ct values. I think they need clear boundaries for low, moderate, and high Ct values with appropriate justification of this categorization.

Of course, describing Ct value in comparative sense is suitable. But even this needs a sort of quantification such as higher in 2 folds or 1 log.

 

4. Discussion:

The discussion, despite being concise, is nicely written. However, a comment needed about some bacterial pathogens, such as C. and M. pneumoniae, that could be present as in carriers.

It is important to mention also the problem of over testing and over trusting these assays as a diagnostic modality.

5. Multiple studies have described interaction between viruses in coinfection (among them PMID: 37722683, 40415261, 37722683, 25943206, 33626106). These studies should be cited and discussed compared to your results.

Author Response

The authors presented an interesting correlation between pathogen detection and pathogen dominance in codetections, in addition to two-year respiratory pathogen prevalence and seasonality.

Actually, it is documented experimentally that influenza virus dominates other pathogens in case of coinfection. Check these two articles (https://doi.org/10.1038/s43856-026-01504-x, https://doi.org/10.1038/s41467-024-53872-4)

We thank the reviewer for highlighting these fundamental studies. We have integrated these citations into the “Discussion” section (page 15, lines 337-338).

The manuscript is interesting to general readers, however, needs to address these comments:

1. Abstract

• The first 2 sentences in the abstract need attention and rephrasing.
• line 18-20: please indicate that the study is retrospective and the years that it spans.

We thank the reviewer for this constructive observation. We have revised the opening of the abstract to correct a grammatical fragment and to explicitly state the retrospective design and the study period (page 1, lines 13-16).

2. Methodology

Study Design and Sample Collection:

• Please mention the years of the data collection to give clear indication of the study period. It is not sufficient to say 2 years.
• This is assay is only validated for nasopharyngeal swabs according to the kit’s leaflet. So nasal swabs are not allowed to be used on this machine in diagnostic settings unless they are validated in the lab internally. Please comment on this.

We thank the reviewer for pointing this out. We have now specified the exact data collection period in the “Study Design and Sample Collection” section (page 3, line 75). Moreover, we sincerely apologize for the confusion caused by a typographical error in the original manuscript. We confirm that only nasopharyngeal swabs were used in this study, fully complying with the manufacturer’s instructions for use. No nasal swabs were collected or analyzed. We have corrected the text in the “Study Design and Sample Collection” section (page 3, lines 76-78).

Ct value extraction and semi-quantitative analysis

• (Lines 99-103) The delta Ct calculation is not quite clear. Please elaborate and clarify

We thank the reviewer for highlighting this lack of clarity. We agree that the original description of the Delta Ct (ΔCt) calculation was ambiguous. We have completely revised the “Ct value extraction and semi-quantitative analysis” section to clarify that the ΔCt is calculated as a pairwise comparison based on the population median Ct values of the co-detected pathogens (page 3, lines 98-104).

Statistical analysis: why did they chose a median based comparison between delta Ct values over the mean delta Ct. This needs to be explained in the manuscript

We thank the reviewer for raising this important methodological point. We chose a median-based approach over a mean-based one because Ct values derived from PCR assays inherently exhibit non-normal, often skewed distributions. Furthermore, these values are bounded by the upper limit of detection of the assay, making the dataset highly susceptible to outliers. In such non-parametric distributions, the mean can be misleadingly skewed by extreme values, whereas the median serves as a much more robust and reliable measure of central tendency. Accordingly, we utilized non-parametric tests (Mann-Whitney U and Wilcoxon signed-rank tests), which evaluate medians and ranks rather than means, to ensure the statistical validity of our comparisons. We have now explicitly added this statistical justification to the “Data Categorization and Statistical Analysis” section of the revised manuscript (page 4, lines 125-129).

 

3. Results

• Table 1 is mentioned before figure 1 in the text but appears after it in the arrangement. Please rearrange.
• Supplementary figure 1 B can be arranged from highest to lowest to give a good idea about the prevalence.
• The same can be done for the table 1

We thank the reviewer for these helpful suggestions to improve the clarity and presentation of our results. We have addressed all three points as follows: Table 1 now appears before Figure 1, consistent with the order of citation in the text, the data in Supplementary Figure 1B have been reordered from highest to lowest prevalence to facilitate visual interpretation and comparison as well as the data in Table 1 have been rearranged in descending order of prevalence to provide a clearer overview of the pathogen distribution.

• Ct values analysis:

The authors described different high and low Ct values with different pathogens. But there is no clear definition of high and low Ct values. I think they need clear boundaries for low, moderate, and high Ct values with appropriate justification of this categorization.

Of course, describing Ct value in comparative sense is suitable. But even this needs a sort of quantification such as higher in 2 folds or 1 log.

We thank the reviewer for this critical observation. We agree that using qualitative terms like “high” and “low” without numerical boundaries, and reporting statistical significance without biological quantification, limits the clinical interpretability of our data. To address this, we have made two major revisions. The first in the “Ct value extraction and semi-quantitative analysis” section (page 3, lines 105-112) defining the boundaries for low (<25), moderate (25–30), and high (>30) Ct values. Later, where we observed statistically significant differences in Ct values between single and multiple detections (specifically for Adenovirus and SARS-CoV-2), we have now calculated and reported the median ΔCt and translated this into the corresponding fold-change in viral load (since PCR amplification is exponential, a ΔCt of 1 corresponds to a 2-fold difference) (page 10-11, lines 219-223; lines 225-228; lines 234-239).

4. Discussion:

The discussion, despite being concise, is nicely written. However, a comment needed about some bacterial pathogens, such as C. and M. pneumoniae, that could be present as in carriers.

We thank the reviewer for this highly relevant clinical observation. We have added a specific comment on C. pneumoniaeand M. pneumoniae in the “Discussion” section (page 15, lines 341-344) to highlight that their detection may reflect a carrier state rather than an active infection.

It is important to mention also the problem of over testing and over trusting these assays as a diagnostic modality.

We completely agree that the high sensitivity of syndromic panels carries the risk of over-testing and of clinicians over-trusting qualitative results without considering the broader clinical picture. We have integrated this concept into the “Discussion” section (page 16, lines 371-378), emphasizing how relying solely on qualitative positive/negative outputs can lead to the misinterpretation of “bystander” pathogens as primary drivers.

5. Multiple studies have described interaction between viruses in coinfection (among them PMID: 37722683, 40415261, 37722683, 25943206, 33626106). These studies should be cited and discussed compared to your results.

We thank the reviewer for highlighting these fundamental studies on viral interactions. We have carefully reviewed the suggested literature and integrated these citations into the “Discussion” section (page 15, lines 347-358).

 

Reviewer 2 Report

Comments and Suggestions for Authors

This retrospective study of 2,479 respiratory samples showed higher pathogen positivity and co-detection rates in children. RSV and influenza demonstrated consistently lower Ct values, suggesting dominant pathogenic roles, whereas RV/EV, adenovirus, and HBoV frequently showed higher Ct values, supporting possible bystander roles. The findings support semi-quantitative interpretation of syndromic panels.

1. The study is retrospective single centre, which may introduce selection bias and limit the generalizability of the findings to other hospitals or populations. The authors should explain how it impacts on the finding or state in its limitation.
2. The interpretation of Ct values as indicators of pathogen “dominance” should be made cautiously because Ct values can vary depending on assay performance, sample quality, and amplification efficiency, and are not equivalent to standardized viral load measurements. 
3. The authors should consider counterchecking the positive cases with their clinical symptoms, severity, radiological findings, treatment, or patient outcomes and then confirm whether the pathogen with the lower Ct value was truly the main causative agent. 
4. The authors should consider adjusting the potential confounding factors such as underlying diseases, immune status, vaccination history, prior antimicrobial use, or duration from symptom onset to sampling, all of which could influence Ct values and pathogen detection. 
5. Some pathogen combinations had very small sample sizes, including triple and quadruple co-detections, which may weaken the statistical reliability of the subgroup analyses and heatmap interpretation. 
6. In table-2, what is meant by HPIV+HPIV, are they different pathogens? Please clarify it and the table heading should be on the top of the table.
7. The conclusion may be slightly overstated because the study only demonstrates associations between Ct patterns and co-detections but does not provide direct evidence that semi-quantitative reporting improves antimicrobial stewardship or clinical outcomes in practice. Prospective validation studies are still needed. 

Author Response

This retrospective study of 2,479 respiratory samples showed higher pathogen positivity and co-detection rates in children. RSV and influenza demonstrated consistently lower Ct values, suggesting dominant pathogenic roles, whereas RV/EV, adenovirus, and HBoV frequently showed higher Ct values, supporting possible bystander roles. The findings support semi-quantitative interpretation of syndromic panels.

1. The study is retrospective single centre, which may introduce selection bias and limit the generalizability of the findings to other hospitals or populations. The authors should explain how it impacts on the finding or state in its limitation.

We thank the reviewer for highlighting this important methodological point. We have explicitly added the retrospective, single-center design and its impact on generalizability in the “Discussion” section (page 16, lines 379-381).

1. The interpretation of Ct values as indicators of pathogen “dominance” should be made cautiously because Ct values can vary depending on assay performance, sample quality, and amplification efficiency, and are not equivalent to standardized viral load measurements. 

We completely agree with this important technical caveat. We have revised the limitations section to explicitly state that Ct values must be interpreted cautiously, as they are influenced by assay performance, sample quality, and amplification efficiency, and are not strictly equivalent to standardized quantitative viral load measurements (page 16, lines 381-384).

1. The authors should consider counterchecking the positive cases with their clinical symptoms, severity, radiological findings, treatment, or patient outcomes and then confirm whether the pathogen with the lower Ct value was truly the main causative agent. 

We agree that correlating Ct values with clinical outcomes is the goal of this approach. However, due to the retrospective, laboratory-based design of this study, detailed clinical metadata (symptom severity, radiology, outcomes) were not available for the 2,479 samples analyzed. We have now clearly stated this lack of clinical correlation as a major limitation and emphasized that future prospective studies are needed to validate the hierarchy against patient outcomes (page 16, lines 384-387).

1. The authors should consider adjusting the potential confounding factors such as underlying diseases, immune status, vaccination history, prior antimicrobial use, or duration from symptom onset to sampling, all of which could influence Ct values and pathogen detection. 

We acknowledge that these are critical confounding factors. Unfortunately, as this was a retrospective study based on laboratory records, data regarding immune status, vaccination history, prior antimicrobial use, and exact symptom onset were not systematically available for adjustment. We have added this to the limitations section to ensure readers interpret the findings with the appropriate context (page 16, lines 387-390).

1. Some pathogen combinations had very small sample sizes, including triple and quadruple co-detections, which may weaken the statistical reliability of the subgroup analyses and heatmap interpretation. 

We thank the reviewer for this observation. We have added a statement in the “Discussion” section acknowledging that the small sample sizes for triple and quadruple co-detections may limit the statistical reliability of the subgroup analyses for those specific scenarios (page 16, lines 390-392).

1. In table-2, what is meant by HPIV+HPIV, are they different pathogens? Please clarify it and the table heading should be on the top of the table.

We apologize for the lack of clarity. “HPIV+HPIV” indicates the co-detection of different Human parainfluenza virus types (HPIV-3 and HPIV-4) within the same sample. We have added a footnote to Table 2 to clarify this. Furthermore, we have moved the table heading to the top of Table 2 as requested.

1. The conclusion may be slightly overstated because the study only demonstrates associations between Ct patterns and co-detections but does not provide direct evidence that semi-quantitative reporting improves antimicrobial stewardship or clinical outcomes in practice. Prospective validation studies are still needed. 

We agree that our study demonstrates associations rather than direct clinical outcomes. We have toned down the conclusion to reflect this, explicitly stating that while prospective validation is still needed to prove direct improvements in antimicrobial stewardship, semi-quantitative reporting has the potential to support diagnostic stewardship (page 16, lines 396-398; lines 400-410).